I read with great interest your comments, and those of many other respondents, about SOC(system on a chip). In particular, you mentioned that National Semiconductor would ultimately need .18 micron technology to add memory(you didn't specify either DRAM or SRAM, but likely both kinds) to the SOC. I come from a processing background with Bell Labs several decades forward, and want to caution investor types that Intel today is trying to shift all its plants to .25 micron, IBM is running .25 micron(and IDTI's C6+ will be run by IBM on it) at their foundry. IDTI's Oregon plant is running .35 micron technology, and Cypress is apparently still running .65 micron technology on its SRAMs. You also might note that AMD has been trying for some time to get its K-6 to yield on .25 micron technology, and only lately has had much luck. It ain't easy! In theory, the concept is to run finer line widths and get smaller die sizes and more die per wafer. That way, the POTENTIAL yield is far better. But if you don't get many(or any) good die with the smaller technology, you are in the tank where AMD has been for the past year. The usable number of die on .35 technology could be larger than those from the same diameter wafer using .25 technology: the ones that don't work can't be sold. I know, I've tried to do it.
To feel that National Semiconductor can buy, install equipment, and manufacture with .18 micron technology when neither IBM nor Intel has done it is blowing smoke. What I'm trying to point out is a system on a chip is a wonderful concept, but may be very hard to reduce to practice. Don't count on it soon. You can tell an analyst and quote me. Frankly, I'll bet even Andy Grove would agree with me on this one. Good luck out there, but I see too many fairy tales on P/E ratios and prospective revenues to pass this up. IDTI is well-positioned in both communications and microprocessor areas, has diversified out of their dependence on SRAM, and their earnings should show it over the next four quarters. I'd hate to see them run into SOC prematurely, and toss the present advantage away. |
